Over the last 15 years, we have had great success in following a sample of formerly psychiatrically hospitalized adolescents (n=180) through adulthood to help us better understand the risk and course of suicidal thoughts and behavior. Using longitudinal data from this study, we have recently examined sensitization as a process that may contribute to recurrent suicidal behavior, the predictive validity of clinical characterisics of suicidal behavior, and differing developmental trajectories of suicide ideation and attempts from adolescence through adulthood. With this revised renewal application, we propose a behavioral model of the recurrence or maintenance of suicidal thoughts and behavior. Four cognitive and affective mechanisms are hypothesized to reflect underlying behavioral processes: (a) delay discounting (a decision-making bias characterized by greater focus on immediate rewards rather than delayed consequences, assessed with the Iowa Gambling Task), and (b) implicit affect (assessed with the Affect Misattribution Procedure), (c) eye blink startle reflex (a physiological indicant of defensive responding), and (d) postauricular reflex (a physiological indicant of approach or appetitive responding) in response to suicide-related stimuli. Using an approach consistent with the Research Domain Criteria (RDoC) initiative of NIH and cognitive and affective neuroscience, we propose to examine these four mechanisms in relation to severity of suicidal thoughts and behavior over time in this longitudinal sample. Th first specific aim of this grant is examine these four mechanisms in relation to severity of suicidl thoughts and behavior from adolescence through young adulthood (i.e., in relation to longitudinal data from 15 years of follow-up). The second aim of this grant is to examine the stability of the mechanisms over time, and the degree to which these variables covary with severity of depression and substance use, as well as psychiatric disorders. The third aim of the grant is to examine the utility of these mechanisms in predicting severity of subsequent suicide ideation and behavior over an additional three years of follow-up. Exploratory aims of this study are to explore (a) the degree to which these cognitive and affective mechanisms are associated with established risk factors for suicidal thoughts/behavior, and (b) the degree to which possible patterns of developmental mediation leading to suicidal ideation/attempts are associated with the hypothesized cognitive and affective mechanisms. For all Aims, we will explore gender and ethnicity effects and the possibility of gender and ethnicity moderation. This study also has clear translational implications for prediction of suicidal thoughts and behavior, and for intervention development for individuals with differing histories of suicidal thoughts and behavior.